TGF-fi tumor suppressor function. However, to date the role of the TGF-/3 pathway at specific stages in gastrointestinal (Gl) tumor development such as metaplasia, dysplasia and carcinoma remains poorly delineated, particularly in conjunction with activation of oncogenic pathways. We previously found that deletion of ELF results in a dramatic and spontaneous formation of liver and gastrointestinal (Gl) cancers, with a splice site mutation in elf exon 15 in 11% of human Gl cancer cell lines tested so far. A surprising and serendipitous recent discovery by us is that elf'and elf^/SmadS^'mice develop visceromegaly and multiple Gl cancers (70% of mice), including metastatic pancreatic, hepatocellular, intestinal adenocarcinomas and others spontaneously, providing compelling evidence as a mouse model of Beckwith-Wiedemann syndrome (BWS), a hereditary human cancer syndrome. In addition, 90% of elf''' /Smad4*/~ mice develop gastric cancer, 20% develop colonic adenomas, and ELF expression is lost in human gastric cancer as well as Dukes B1 adenomas indicating a role for ELF in suppression of early human gastrointestinal cancer. Molecular profiling of the tumors in these mice and human Gl cancers demonstrate markedly high levels of cell cycle regulators that include CDK4, cyclin D1 and PRAJA an ELF/Smad3 specific E3 ligase. Interestingly, SmadS has recently been shown to be a CDK4 substrate, yet Smad3 mutant mice do not develop cancers. Thus, activated CDK4 and PRAJA associate with ELF with or without Smad3 and most likely exert their oncogenic activity through suppression ofELF/SmadS. Our AIMS are to: 1. Determine a functional interaction between ELF, SmadS, and CDK4, and the effect of loss elf, SmadS or Smad4 on CDK4 enzyme activity, towards testing CDK4 inhibitors from project 4, Vitamin D analogs from project 2, as well as developing new therapaeutics targeted allosterically at CDK4 -ELF-Smad3 interaction. 2. Investigate the molecular basis for the differential effects of PRAJA on ELF and SmadS towards generating new inhibitors targeted at PRAJA. 3. Test whether ELF is mutated in BWS, and potentially whether loss of ELF in combination with PRAJA, CDK4, Smad3, Smad4,TBRII, TERT and c-Myc (the latter two from project 3) represent new specific molecular markers for early tumor detection/ treatment response of hepatocellular, gastric and pancreatic cancers. Results from this study promise to yield important new therapeutics and will be a first step toward the goal of individualized cancer treatment based on the functional molecular characteristics of these lethal tumors.